时空数据Web搜索的研究进展

Research Progress on Searching Spatiotemporal Data from Web

随着移动Web和互联网模式的逐步兴起,时空大数据应运而生,如何从泛在网络中挖掘可用资源成为各界广泛关注的热点问题。而Web搜索是从互联网中精准发现并提取时空大数据的基本手段。以经典Web搜索为理论基础,针对时空数据的时空特征和多源多模态特征,分析了由此带来的空间语义特殊化、时间意图明确化和时空数据对象化3个方面的基本问题,并以此评述相关研究进展。最后,总结并展望了时空数据深层Web自适应采集、多源多标签遥感图像检索、跨模态时空数据检索、时空数据Web搜索工程化实施等方面的发展方向。

With the gradual rise of mobile Web and Internet models,spatiotemporal big data has increased quickly.How to mine the available resources from the ubiquitous web becomes a hot topic.Web search is the basic means of extracting spatiotemporal big data from the Internet.Based on the classical web search theory,this paper analyzes three basic problems,i.e.spatial semantic specialization,temporal intent clarification,and temporal data objectification of spatiotemporal data.Then,this paper reviews the progress of relevant research.Finally,the future research directions of web search for spatiotemporal data are briefly summarized in terms of deep web crawling,multi-source and multi-label remote sensing image retrieval,cross-modal retrieval,and web search implementation.